Adrien Giget

Expert in Generative AI, NLP, and Machine Learning, specializing in Agentic AI systems and modular RAG pipelines. I design and deploy end-to-end full-stack AI solutions to transform raw data into strategic insights.

With completed projects including:

- AI agents that automatically extract data from any supplier document (PDF, Excel, unstructured formats) and convert it to ERP format. Extraction accuracy reaches nearly 100%, validated on an evaluation framework with reference datasets and per-field metrics.

- Development of a SaaS platform dedicated to the statistical analysis of French professional volleyball. End-to-end solution transforming raw match data into strategic insights for coaches, statisticians, and players via interactive dashboards and predictive modeling.

- Development of an internal AI platform at the client's site to analyze text modules into reasoned and sourced responses. Enabling the client to quickly position themselves on the answers given in forms.

- Development of a chatbot to lighten the administrative workload by automatically answering student questions while respecting Trinity College Dublin's internal liabilities.

- Use of open-source tools and Azure/OpenAI to create a high-performance RAG pipeline at the state-of-the-art level.

- Cross-functional use of the project, both for its primary need and transversally to audit the data present on the different parts of the TCD website (which manages 3 faculties and 24 schools) to standardize internal documents.

- Fine-tuning of BERT and GPT models for incident categorization into tickets, with good final performance (~90%).

- Training of LSTMs and transformers for time series prediction.

- Training of models for predicting incidents and their resolution time.